xml-dev - Re: Again wit da AND and Repetitions

[Lars Marius Garshol <larsga@ifi.uio.no>]

* roddey@us.ibm.com
|
| I never really had any one say whether they thought that validation
| that included both AND connectors and M to N repetition for any of
| the connectors is feasible (meaning fast and compact and
| maintainable and verifiable in my opinion, but tell me if you
| disagree.)
| 
| Does anyone believe that these can be done in less than brute force
| tree search times?

I think so, yes. From skimming through the draft once and thinking a
little about it I think it will cost noticeably more than normal
validation, though. At least for my parser it seems that it will.

My best idea so far is to create tracker objects that move through the
content model as you parse the document. This is just a sketch in my
head, though, and nothing is written or coded yet. (I'm currently
working on support for datatypes and haven't gotten round to the
structural schemas yet.)

* Terje Norderhaug
| 
| XML requires deterministic content models. This allows validators to
| do their job without having to look more than one element ahead or
| do brute force tree searches.

Determinism doesn't have anything to do with it. You can still easily
create deterministic finite state automata for XML content models (as
described in DTDs, that is). Since the tags in XML remove any chances
of ambiguity this doesn't cause any ambiguity in interpreting the
markup. In fact, I need to add special code to my parser in order to
warn about non-deterministic ones.

In theory, you can do the same for what is described in the XML
schemas specification, but in many cases you'll then be looking at a
combinatorial explosion of automaton states. In fact, with DTDs like
the XTEILITE DTD things can get pretty bad as it is.

--Lars M.


xml-dev: A list for W3C XML Developers. To post, mailto:xml-dev@ic.ac.uk
Archived as: http://www.lists.ic.ac.uk/hypermail/xml-dev/ and on CD-ROM/ISBN 981-02-3594-1
To (un)subscribe, mailto:majordomo@ic.ac.uk the following message;
(un)subscribe xml-dev
To subscribe to the digests, mailto:majordomo@ic.ac.uk the following message;
subscribe xml-dev-digest
List coordinator, Henry Rzepa (mailto:rzepa@ic.ac.uk)